PolaCalibration - Calibration of Polarization Cameras

A Practical Calibration Method for RGB Micro-Grid Polarimetric Cameras
[Paper arXiv] [Code] [Supplem Material]

This repository contains the original implementation of the polimetric calibration paper "A practical calibration method for RGB micro-grid polarimetric cameras", submitted to RA-L 2022.

(a) Intensity image	(b) Uncalibrated AoLP	(c) Calibrated AoLP

Our calibration approach finds the parameters of camera and lens to generate the super-pixel matrix, as shown in the results of the calibration for the angle of linear polarization (AoLP). (a) RGB Intensity image. (b) Uncalibrated AoLP. (c) Calibrated AoLP. The AoLP has been colored using the HSV color palette to allow the circularity of the colors (180 and 0 degrees should be represented with the same color).

Citation

If you find this code useful for your research, please cite the paper:

@ARTICLE{rodriguez-ral2022,
  author={Rodriguez, Joaquin and Lew-Yan-Voon, Lew and Martins, Renato and Morel, Olivier},
  journal={IEEE Robotics and Automation Letters}, 
  title={A Practical Calibration Method for RGB Micro-Grid Polarimetric Cameras}, 
  year={2022},
  volume={7},
  number={4},
  pages={9921-9928},
  doi={10.1109/LRA.2022.3192655}}

Code

In this folder, the paper implementation code is included. It consists of C++ code with the developed method, and some raw images to test its performance.

Supplementary documents

In this folder, we detail some points which due to space reasons, could not be included in the main paper:

1. The theoretical demonstration about the angle of polarization error.
2. The results of using a calibrated camera in real robotics applications.

Nevertheless, these explanations are not part of the paper. The main text stands by itself. These explanations are addressed to the readers who want to know the additional details of the calibration modelling presented in the main paper.

Supplementary images

In this folder, images that could not be included in the paper are attached. Since the RGB channels from the image are considered as three separated images, then there are three different calibration results per frame. We have presented the red channel results in the paper, while in this folder includes the analysis for the red, green and blue channels.

Additionally, two set of images have been added, which correspond to the high resolution version of the applications result images.